Exploring Grover-Enhanced QAOA for Minimal Cut Set Identification

• Published in: Proceedings. Annual Reliability and Maintainability Symposium pp. 1 - 6
• Main Authors: Silva, Gabriel San Martin, Droguett, Enrique Lopez
• Format: Conference Proceeding
• Language: English
• Published: IEEE 27.01.2025
• Subjects: Computational complexity; Computational modeling; Fault diagnosis; Fault Tree Analysis; Fault trees; Grover; Hardware; Minimal Cut Set Identification; Mixers; Optimization; QAOA; Quantum Computation; Quantum computing; Random access memory
• ISSN: 2577-0993
• DOI: 10.1109/RAMS48127.2025.10935088

Fault Tree Analysis is a powerful technique used in a wide variety of fields to perform both quantitative and qualitative analysis of complex systems. Among the capabilities of Fault Tree Analysis, one of the most critical and computationally expensive is the identification of Minimal Cut Sets. With the ever-growing demand for more efficient, sustainable, and safe systems, we can expect the computational complexity of finding minimal cut set configurations to increase even further. To prepare for this challenge, this paper explores the feasibility of finding minimal cut set configurations in standard, coherent fault trees using a novel quantum-based optimization approach. The approach, previously proposed to find satisfying clauses for 3-SAT problems in general settings, is modified to accommodate the context of Fault Tree Analysis. To validate the proposed approach, we perform numerical tests on a quantum computing simulation environment. The results show that while the identification of minimal cut sets is feasible, challenges arise when the size of the fault tree increases. Overall, the experimental results verify the feasibility of identifying minimal cut sets in standard, coherent fault trees using the G-QAOA methodology. However, due to hardware limitations, this verification was only conducted in relatively small fault tree models. Regarding the use of a modified mixer operation, as proposed by Mandl et al., the results demonstrate significant performance advantages over the mixer operator used in traditional QAOA. Therefore, our main recommendation for future studies in this area is to focus on incorporating this modified operator into newly developed models.